ES2980627T3 - Respiratory rate measuring device - Google Patents

Abstract

Respiratory rate measuring device (4) used for a PSA type oxygen concentration device and capable of improving measurement accuracy, characterized in that the measuring device comprises a detection unit (6) that is connected to a patient and that detects an intratubal pressure and/or an intratubal air flow rate in a tube (2) that supplies concentrated oxygen gas to the patient from a pressure swing adsorption type oxygen concentration device (1) that concentrates oxygen in the air by periodically repeating pressurization and depressurization, and that generates pressure data and/or air flow rate data, a calculation unit (722) that extracts patient respiration information data based on the pressure data and/or the air flow rate data, and an estimation unit (723) that estimates a prescribed respiration rate per unit time based on the patient respiration information data, where the estimation unit (723) derives an autocorrelation coefficient between the patient respiration information data and the patient respiration information data. which are displaced relative to the patient's breathing information data by a time Δt. while adjusting said time Δt, and estimates the breathing rate using the time Δt at which the autocorrelation coefficient reaches a peak as the breathing interval. (Automatic translation with Google Translate, no legal value)

Description

DESCRIPTION

Respiratory rate measuring device

FIELD

The present disclosure relates to a respiratory rate measuring device used for a pressure swing adsorption oxygen concentration device.

BACKGROUND

Oxygen therapy has been conventionally performed as a medical treatment for patients suffering from respiratory diseases such as asthma and chronic obstructive pulmonary disease. In this treatment, patients inhale oxygen gas or concentrated oxygen gas. In recent years, OTD (home oxygen therapy) in which patients inhale oxygen, for example, at home or in facilities, with the aim of improving patients' QOL (quality of life), is becoming mainstream, and an oxygen concentrating device is mainly used as the oxygen source.

Oxygen concentrating device means a device that concentrates and supplies oxygen existing at about 21% in the air. Most oxygen concentrating devices are generally of the pressure swing adsorption type (hereinafter referred to as the PSA type).

The PSA oxygen concentrating device repeatedly performs an adsorption process of introducing air into an adsorption column filled with an adsorbent to selectively adsorb nitrogen gas, and pressurizing the adsorption column to adsorb the nitrogen gas to the adsorbent, and a desorption process of depressurizing the adsorption column to expel the adsorbed nitrogen gas from the system. By repeating the adsorption and desorption processes, concentrated oxygen gas is generated to enable the oxygen concentrating device to continuously provide highly concentrated oxygen gas to the patient. A method of depressurizing the adsorption column to atmospheric pressure or less in the desorption process is sometimes not called PSA but, for example, VPSA or VSA, but the basic principle of VPSA or VSA is the same as that of PSA, and these methods will be collectively referred to hereinafter as PSA.

The main disease of patients receiving home oxygen therapy is chronic obstructive pulmonary disease (hereinafter referred to as COPD). COPD refers to an irreversible disease that produces symptoms of cough and expectoration of sputum or dyspnea when exercising, due to constriction of the bronchi or destruction of the alveolar walls of the lungs.

When COPD symptoms worsen, shortness of breath or increased respiratory rate is observed, and clinical manifestations sometimes deteriorate until the “state where a change of treatment or any additional treatment may be preferentially involved in a stable phase” called acute exacerbation of COPD occurs. When such acute exacerbation of COPD occurs, patients are admitted to hospitals in most cases and may even go into respiratory failure or face crises of their lives. Moreover, even when patients are allowed to leave hospitals, it is often the case that patients develop more severe symptoms in the stable phase than they had before their hospital stay, and are then repeatedly admitted and discharged from hospitals.

In COPD, it is very important to detect early signs or symptoms of an acute exacerbation in the patient as early as possible, and to treat the patient soon before his/her clinical manifestations worsen to the extent that he/she is hospitalized, and in home oxygen therapy, therefore, the patient's respiratory information, especially a change in the patient's respiratory rate, serves as a very beneficial source of information in terms of knowing the patient's pathological condition.

Respiratory rate can be measured even by using, for example, a stomach belt, but in order to find out about a long-term change in the patient's pathological condition, the patient is forced to bear a load, since, for example, he can preferably wear the stomach belt at all times, or periodically taking into account its use at all times. Generally, patients receiving home oxygen therapy are expected to periodically wear the stomach belt for at least several hours a day. As long as a respiratory rate measuring device is built into or attached to an oxygen concentration device, the respiratory rate measuring device can reliably check for a change in the pathological condition during use of the oxygen concentration device, and without forcing the patient to bear an additional load. The respiratory rate measuring device is very useful.

As a method for obtaining a respiratory rate of a patient, which may be implemented in an oxygen concentrating device, there is available a method for mounting a microdifferential pressure sensor for respiratory measurement between the oxygen concentrating device and a cannula mounted on the patient to measure a respiratory pressure of the patient during oxygen inhalation, and recording the respiratory pressure of the patient on a recording medium or transmitting it as communication data, as described in PTL 1.

PTL documents 2 and 4 disclose the fact that respiratory rate and the like can be calculated from a respiratory pattern, but do not describe specific methods.

Document PTL 3 discloses a method for storing the times at which the pressure waveform changes from a fall to a rise and counting the respiration rate, based on the interval between these times, and a method for multiplying the pressure amplitude by a predetermined detection level ratio and determining that respiration is present only when a threshold corresponding to the calculated product is exceeded.

PTL 5 describes a method for calculating respiratory rate using FFT (fast Fourier transform) processing or TDS processing.

PTL 6 describes a method for measuring and storing pressure variations of an oxygen concentration device itself in advance, and subtracting the pressure variations from a detected pressure waveform.

PTL 7 describes an oxygen concentrator that takes advantage of pressure swing adsorption. A flow sensor is used to determine the flow rate of gas flowing through the outlet system to the user. A pressure sensor is used to monitor the pressure of the gas passing through the line to the user. A controller collects and stores the number of breaths. Based on the number of breaths, or the average time between breaths, over the time period a respiratory rate can be calculated.

PTL 8 discloses the analysis of a photoplethysmography (PPG) signal using an autocorrelation sequence. The time between peaks of the combined autocorrelation sequence that are representative of respiration information may be equivalent to the period of respiration, which may be used to determine the respiratory rate (e.g., breathing rate).

[QUOTE LIST]

[PATENTS BIBLIOGRAPHY]

[PTL 1] JP H6-190045 A

[PTL 2] JP 2001-286566 A

[PTL 3] JP H7-96035 A

[PTL 4] JP 2015-85191 A

[PTL 5] JP 2011-518016 A

[PTL 6] WO 2018-180392 A1

[PTL 7] US 2016/375215 A1

[PTL 8] US 2015/230759 A1

SUMMARY

In any method of directly sending the calculated respiratory data, since the waveform data measured every 100 milliseconds (ms), for example, is transmitted directly, the amount of data is enormous, and it may take a long time for analysis.

The method of directly outputting the calculated respiratory data has the drawback that, when an irregularity occurs in the respiratory waveform due to, for example, noise or body movement, a peak value or the like generated by the irregularity is detected as breathing. The respiratory waveform changes considerably depending on the pathological condition of a patient wearing the respiratory rate measuring device, or the patient's state such as an active state or a sleep state, and even varies considerably in each individual patient. It is difficult to set a threshold as a criterion for breathing under a uniform rule, and to determine whether breathing is present.

In FFT, a huge amount of calculation is involved, and furthermore, when the data interval used for arithmetic operation is set short, no accurate respiratory period can be calculated, and when the data interval is set too long, hardly a peak of respiratory period can be obtained due to the influence of a slight change in respiratory interval for each breath. TDS processing poses a problem related to the influence of noise or body motion, a change in respiration depending on the patient's condition, and differences in respiratory status for each individual patient.

A respiratory rate measuring device has been invented to solve the problems described above, and has as its exemplary object to make it possible to improve the accuracy of respiratory rate measurement, using a PSA oxygen concentration device in combination.

This object is achieved by the subject matter of the independent claim. The dependent claims relate to particular embodiments.

According to this embodiment, the respiratory rate measuring device can improve the accuracy of respiratory rate measurement, without forcing the patient to bear additional burden, by using a PSA oxygen concentration device in combination.

The objects and effects of the present invention will be appreciated and obtained by means of the elements and combinations particularly pointed out in the appended claims. Both the foregoing general description and the following detailed description are exemplary and explanatory, and do not limit the present invention as described within the scope of the claims.

Brief description of the drawings

Fig. 1 is a diagram illustrating an exemplary schematic configuration of a PSA oxygen concentrating device.

Fig. 2 is a diagram illustrating an exemplary schematic configuration of a respiratory rate measuring device.

Fig. 3 is a block diagram illustrating an exemplary microcomputer unit.

Fig. 4 is a graph depicting pressure data including patient respiratory information, under the conditions where the continuous flow is set to 5 LPM, and an extension tube is added in 20 m.

Fig. 5 is a diagram representing PSA pressure data extracted by an arithmetic operation unit, under the conditions that the continuous flow is set to 5 LPM, and an extension tube is added at 20 m. Fig. 6 is a graph representing the patient's respiratory information after subtraction processing, under the conditions that the continuous flow is set to 5 LPM, and an extension tube is added at 20 m.

Fig. 7 is a conceptual diagram depicting the principle of separating patient respiratory information and PSA pressure from pressure data that includes pressure and PSA pressure variations caused by the patient's breathing.

Fig. 8 is a graph representing the result of calculating an autocorrelation coefficient from the patient's respiratory information after subtraction processing.

Fig. 9 is a flowchart illustrating an exemplary processing of patient respiratory information data extraction.

Fig. 10 is a flowchart illustrating an exemplary processing of respiratory rate estimation, based on patient respiratory information data.

Description of realizations

A respiratory rate measuring device according to one aspect of an embodiment or embodiments will now be described with reference to the drawings. However, it should be noted that the technical scope of the present disclosure is not limited to these embodiments and encompasses the invention described within the scope of the claims. In the following description and drawings, the same reference numerals denote components having the same functional configurations, and a repetitive description thereof will not be given.

<PSA oxygen concentration device>

The configuration of a PSA oxygen concentrating device, implemented as an exemplary pressure swing adsorption oxygen concentrating device, used in combination with a respiratory rate measuring device in accordance with one aspect of an embodiment will now be described. Fig. 1 is a diagram illustrating an exemplary schematic configuration of a PSA oxygen concentrating device.

A PSA oxygen concentration device 1 includes an oxygen generating unit 11 that generates concentrated oxygen gas by introducing air A from the outside of the PSA oxygen concentration device 1. The air A introduced from the outside of the PSA oxygen device into the oxygen generating unit 11 is compressed by a compressor 111 and fed to an adsorption column 113 through a first switching valve 112. The first switching valve 112 connects the compressor 111 to one of the adsorption columns 113 to feed the compressed air to the connected adsorption column 113, and opens the remaining adsorption columns to the atmosphere.

The adsorption columns 113 are filled with adsorbents for selectively adsorbing nitrogen gas. Compressed air that has passed through the adsorption column 113 reduces the nitrogen gas concentration and thus becomes concentrated oxygen gas. The concentrated oxygen gas is stored in a concentrated oxygen buffer tank 115 via a second switching valve 114. The second switching valve 114 connects or disconnects one of the adsorption columns 113 to or from the concentrated oxygen buffer tank 115.

The oxygen generating unit 11 uses the first switching valve 112 to connect the compressor 111 to one of the adsorption columns 113, and uses the second switching valve 114 to connect the adsorption column 113 connected to the compressor 111 to the concentrated oxygen buffer tank 115. Therefore, the compressor 111, one of the adsorption columns 113, and the concentrated oxygen buffer tank 115 are connected to each other, and the generated concentrated oxygen gas is supplied to the concentrated oxygen buffer tank 115. The adsorption columns 113 which are not connected to the compressor 111, however, are opened to the atmosphere through the first switching valve 112, since they are disconnected from the concentrated oxygen buffer tank 115 by the second switching valve 114. With this operation, the adsorption columns 113 are depressurized to expel the nitrogen gas adsorbed to the adsorbents out of the PSA oxygen concentration device 1.

The opening and closing of the first switching valve 112 and the second switching valve 114 are controlled, for example, by a microcomputer unit in a respiratory rate measuring device (not illustrated). The microcomputer unit may acquire a timing for switching between pressurization and depressurization in the adsorption columns 113. The respiratory rate measuring device may be placed in the oxygen concentration device 1, or may be placed outside the oxygen concentration device 1 separately from the oxygen concentration device 1.

As another example, the oxygen concentration device 1 may include an oxygen concentration control unit that controls oxygen concentration processing including opening and closing processing of the first and second switching valves. The microcomputer unit may acquire a timing for switching between pressurization and depressurization in the adsorption columns 113 from the oxygen concentration control unit.

The PSA oxygen concentration device 1 may connect two or more adsorption columns of the adsorption columns 113, in addition to the basic configuration mentioned above. The PSA oxygen concentration device 1 may include an additional process such as a pressure equalization process to equalize the pressures in the respective adsorption columns, or a purge process to reflux part of the generated concentrated oxygen gas to one of the adsorption columns 113.

Normally, after compressed air is fed to the adsorption column 113, the adsorption column 113 is opened to the atmosphere as it is disconnected from the compressor 111 by the first switching valve 112. In contrast to this, any adsorption column 113 open to the atmosphere is connected to the compressor 111 by the first switching valve 112, and is shifted to the oxygen compression process. In this manner, use of the first switching valve 112 to allow the adsorption columns 113 to alternately repeat compression and opening to the atmosphere makes it possible to continuously supply concentrated oxygen gas.

Since the pressure change in the adsorption column 113 after generation of concentrated oxygen is very large, periodic pressure variations in the pressure occur within an oxygen gas flow path formed downstream of the adsorption column 113 after switching of the adsorption column 113. The concentrated oxygen gas stored in the concentrated oxygen buffer tank 115 is regulated to attenuate the pressure variations by a pressure regulating valve 116.

The concentrated oxygen gas having its pressure regulated by the oxygen generating unit 11 has its oxygen flow rate controlled by an oxygen flow control unit 12 formed by a control valve 121 and a flowmeter 122, and is supplied to the outside of the oxygen concentrating device through an oxygen supply port 13 via a humidifier 101. In the oxygen flow control unit 12, the control valve 121 or the flowmeter 122 may be provided upstream in the flow path, and the oxygen flow control unit 12 may include other configurations or arrangements.

The PSA oxygen concentrating device 1 may include a fixed switching orifice for switching flow rate, instead of the flow meter 122 and the control valve 121. The PSA oxygen concentrating device 1 may also use a manual flow rate control scheme using a visually observable flow meter such as a rotor meter like the flow meter 122, and a manual flow control valve instead of the control valve 121, or it may use other flow control schemes. The PSA oxygen concentrating device 1 may even have a configuration equipped without a humidifier 101.

<Respiratory rate measuring device>

An exemplary schematic configuration of a respiratory rate measuring device according to the present invention will now be described.

Fig. 2 is a diagram illustrating an exemplary schematic configuration of a respiratory rate measuring device.

Oxygen generated by the PSA oxygen concentrating device 1 is supplied to the nasal cavity of a patient through a tube 2 connected to the PSA oxygen concentrating device 1, and a nasal cannula 3 connected to the tube 2. The patient breathes constantly even during oxygen inhalation, and a pressure change occurring as the patient breathes propagates to the nasal cannula 3, the tube 2, and the PSA oxygen concentrating device 1.

In this embodiment, to obtain the patient's respiratory pressure, a respiratory rate measuring device 4 is connected to the tube 2 serving as an oxygen supply path. The tube refers to one including the entire tube provided between the humidifier 101 (the oxygen flow control unit 12 when the PSA oxygen concentration device 1 is not equipped with the humidifier 101) and the nasal cannula 3, and the respiratory rate measuring device 4 may be connected anywhere in the tube.

Part or all of the respiratory rate measuring device 4 may be placed inside or outside the PSA oxygen concentration device 1.

The inventors of the present invention made a close study and found that even after attenuating the pressure variations by the pressure regulating valve 116, the pressure variations applied to the respiratory rate measuring device 4 include those occurring after pressurization and depressurization during generation of concentrated oxygen gas. Since the amplitude of the pressure variations occurring after generation of concentrated oxygen gas is larger than that of the respiratory pressure, and the amplitude of the respiratory pressure is also reduced due to a pressure loss occurring after passing through the oxygen flow path, it is difficult to directly measure the patient's respiratory pressure waveform from the pressure measured by the respiratory rate measuring device 4.

In this embodiment, the respiratory rate measuring device 4 includes a microcomputer unit 7 including an arithmetic operation unit and an estimation unit and connected to a pressure sensor 6. The pressure sensor 6 is preferably configured as a microdifferential pressure sensor. Furthermore, the respiratory rate measuring device 4 preferably includes a display unit 8, such as a liquid crystal display, which displays the measured respiratory rate. The display unit 8 is connected to the microcomputer unit 7 and controlled by the microcomputer unit 7.

In this embodiment, the respiratory rate measuring device 4 includes, for example, a pressure sensor 6, and preferably a micro-differential pressure sensor 6, as a detection unit that detects and outputs the pressure in the tube, and a microcomputer unit 7 electrically connected to the detection unit. The respiratory rate measuring device 4 may further include a positive displacement unit connected to the micro-differential pressure sensor 6, and a pressure smoothing unit formed by a hole 5 connecting the tube and the positive displacement unit to each other. The reason why the pressure smoothing unit is provided is as follows.

Since the patient's respiratory pressure is normally about ±10 to 100 Pa, a sensor having the range of about ±100 Pa is preferably used as the microdifferential pressure sensor 6 to obtain the respiratory pressure by the respiratory rate measuring device 4. In the state where oxygen is supplied from the oxygen concentration device, a supply pressure is constantly generated after oxygen supply, and the supply pressure generated after oxygen supply exists at, for example, about 300 Pa even for 1 LPM (stretcher per minute: l/min). This means that when one end of the microdifferential pressure sensor 6 is connected to the oxygen supply path, as described above, with the other end of the microdifferential pressure sensor open to the atmosphere, the pressure to be measured falls outside the measurement range of the microdifferential pressure sensor 6. Therefore, a pressure including respiratory information of the patient in the measuring range of the microdifferential pressure sensor 6 is preferably obtained by applying the pressure after passing through the hole 5 to the other end of the microdifferential pressure sensor 6.

A pressure falling within the measurement range of the microdifferential pressure sensor 6 may preferably be applied to the other end of the microdifferential pressure sensor 6, and a method for this operation is not limited to the example in this embodiment. As long as the measurement range of the microdifferential pressure sensor 6 is larger than the supply pressure generated after oxygen supply, and a high enough resolution can be ensured to enable detection of pressure variations occurring after a patient's breathing, the other end of the microdifferential pressure sensor 6 may even be opened to the atmosphere. A pressure sensor measuring a gauge pressure or an absolute pressure may be replaced by the microdifferential pressure sensor 6.

Since a pressure change during breathing also appears as minimal variations in the flow rate of the concentrated oxygen gas flowing through the tube, the respiratory rate measuring device 4 can use a flow sensor instead of the pressure sensor 6. The respiratory rate measuring device 4 can even include both the pressure sensor 6 and the flow sensor. The pressure sensor 6 and/or the flow sensor serve as a sensing unit which detects a pressure in the tube and/or a gas flow rate in the tube 2 including respiratory information of the patient, and outputs pressure data and/or gas flow rate data in the tube.

<Microcomputer unit>

Figure 3 is a diagram illustrating an exemplary configuration block of the microcomputer unit 7.

The microcomputer unit 7 including an arithmetic operation unit 722 and an estimation unit 723 is connected to the pressure sensor 6, and preferably to the microdifferential pressure sensor 6. The arithmetic operation unit 722 and the estimation unit 723 obtain respiratory information of the patient by receiving the in-tube pressure data and or the in-tube gas flow rate data in the tube including the respiratory information of the patient, detected by the microdifferential pressure sensor 6 serving as the detection unit, and performing the processing involved (to be described later).

The microcomputer unit 7 may be implemented as the same microcomputer as in a processing unit that processes, for example, an oxygen generation function and a display and user interface function for the oxygen concentration device, or it may be separate from the processing unit. When the microcomputer unit 7 is separate from the processing unit, it acquires a timing for switching a PSA period T from a microcomputer that processes the oxygen generation function, and uses the acquired timing for the arithmetic operation.

The microcomputer unit 7 includes a storage unit 71 and a processing unit 72. The storage unit 71 is configured as one or more semiconductor memories. The storage unit 71 includes at least one of non-volatile memories, such as a RAM, a flash memory, an EPROM and an EEPROm. The storage unit 71 stores, for example, a driver program, an operating system program, an application program and data used for processing by the processing unit 72.

The storage unit 71 stores, as a driver program, for example, a device driver program for controlling, for example, the micro-differential pressure sensor 6 serving as a detection unit. A computer program can be installed on the storage unit using, for example, a known configuration program of a portable computer-readable recording medium such as a CD-ROM or a DVD-ROM. The computer program can even be downloaded from, for example, a program server and installed.

The storage unit 71 may further temporarily store temporary data associated with a predetermined processing. The storage unit 71 stores, for example, a threshold 711 and a variation value data file 712 for use in estimating the respiratory rate.

The processing unit 72 includes one or more processors and their peripheral circuits. The processing unit 72 systematically controls the overall operation of the respiratory rate measuring device 4, and is implemented as, for example, an MCU (microcontrol unit).

The processing unit 72 performs processing based on the programs (for example, the operating system program, the driver program, and the application program) stored in the storage unit 71. The processing unit 72 may execute multiple programs (for example, the application program) in parallel. The processing unit 72 includes, for example, a sensed data acquisition unit 721, the arithmetic operation unit 722, the estimation unit 723, and a respiratory rate output unit 724.

Each of these units constituting the processing unit 72 may be implemented in the microcomputer unit 7 as an independent integrated circuit, a circuit module, a microprocessor or a firmware.

<Principle of arithmetic processing performed in the implementation>

The processing principle performed in this embodiment, using respiratory information data obtained by applying a respiratory pressure on a patient respiratory model from the nasal cannula 3 using the configuration illustrated in Figure 2, will be described below.

Figures 4, 5 and 6 illustrate a set of data obtained when the continuous flow is set to 5 LPM, and a 20 m extension tube is connected on the downstream side of the respiratory rate measuring device 4. Continuous flow means a concentrated oxygen gas supply scheme, in which concentrated oxygen gas is continuously supplied at a constant flow rate.

Fig. 4 is a graph depicting pressure data including patient respiratory information and PSA pressure data of the PSA oxygen concentration device 1, when the continuous flow is set to 5 LPM, and a 20 m extension tube is connected.

Fig. 5 is a graph depicting the PSA pressure data of the PSA oxygen concentration device 1 if the patient is not breathing or after the respiratory components are removed, when the continuous flow is set to 5 LPM, and a 20 m extension tube is connected.

The PSA pressure obtained by the PSA oxygen concentration device means a periodic pressure change related to the period of the adsorption column of the PSA oxygen concentration device 1 and which occurs when oxygen is generated by the PSA oxygen concentration device 1, as described above. The period of the PSA pressure waveform coincides with the switching period of the adsorption column of the PSA oxygen concentration device 1.

Fig. 6 is a graph representing the result of subtraction processing of the components illustrated in Fig. 5 from the components illustrated in Fig. 4 by software. Subtraction processing means the processing of calculating the difference between two pieces of data at an arbitrary time instant. By subtraction processing, the respiratory rate measuring device 4 can remove PSA pressure components and detect a respiratory pressure in a patient respiratory model. By performing subtraction processing by software, the respiratory rate measuring device 4 can obtain respiratory information of the patient even in the state where oxygen is continuously inhaled at 5 LPM from the PSA oxygen concentration device 1 through the nasal cannula and the 20 m extension tube.

As a method for extracting respiratory information data of the patient by removing variation value data such as PSA pressure components, the method for measuring and storing pressure variations of an oxygen concentration device itself in advance, and subtracting the pressure variations from a detected pressure waveform, as disclosed in PTL 6, may be used, or a method for subtracting pressure variations of an oxygen concentration device itself measured in real time from a detected pressure waveform may be used.

When the PSA oxygen concentrating device switches between pressurization and depressurization with a single period T for at least a part of the total operating time, the arithmetic operation unit 722 extracts respiratory information by calculating, as the pressure at a certain time t, the difference between a flow rate value Y(t) and a value X(t) obtained by averaging Y(t) and Y(t - T), Y(t - 2T),..., Y(t - nT) (n is a preset arbitrary integer). An original respiratory waveform having the PSA pressure components removed can therefore be accurately reproduced.

Fig. 7 is a conceptual diagram depicting arithmetic processing according to this embodiment. (I) illustrates a respiratory waveform, and (II) illustrates a model waveform for the PSA waveform. T denotes the period of the PSA waveform, which coincides with the switching period of the adsorption columns. (III) illustrates the sum of the waveforms illustrated in (I) and (II), and corresponds to the pressure measured in the tube portion. (IV) illustrates the result of dividing the waveform of (III) for each period T and superimposing the divided waveforms on each other. Unless the respiratory period and the PSA period completely coincide with each other, respiratory waveforms randomly appear in the waveforms divided by the periods T. (V) illustrates the result of averaging the waveforms superimposed on each other in (IV). Based on the measured pressure data and/or gas flow rate data (III), variation value data X(t) representing a periodic pressure change and/or a concentrated oxygen gas flow rate change generated by the operation of the PSA oxygen concentration device are estimated by superposition and averaging (V).

In one example, assuming T as a period, a simple moving average of five periods is used. (VI) illustrates a waveform obtained by subtracting the waveform of (V) from the part corresponding to the last period T of the waveform of (III). This reveals that the original respiratory waveform (I) can be accurately reproduced. As an exemplary period T, the time from the start of the state in which one of the adsorption columns is connected to the compressor until a changeover is made to the state in which another adsorption column is connected to the compressor can be selected. As another exemplary period T, the time from the start of the state in which one adsorption column is connected to the compressor, and then through the state in which the remaining adsorption columns are connected to the compressor, until a changeover is made to the state in which the first adsorption column is connected to the compressor can be selected again. As yet another example, a multiple of each switching time can be selected as T.

In this embodiment, when calculating the waveform of (V) representing the variation value data X(t) obtained by superposition and averaging and estimated to carry the information of a periodic pressure change and/or a flow rate change of the concentrated oxygen gas, a five-period moving average is used, but another method may be used for averaging. The average value X(t) calculated in the step of (V) is generally given by the following equation:

[Eq. 1]

W_ £ jU

goes

y1in^lg g z g

=0^ .(1)

Throughout the description of this disclosure, i does not represent an imaginary number, but simply represents an integer variable i.

X(t) is the value after average processing at time t, and Y(t) is the actual measured value of pressure at time t i is a weighting coefficient in weighted averaging, which can be obtained by selecting an arbitrary real number, n is the number of Y values to be averaged. For, for example, n = 4, and a<0>a a4 = 1, the same calculation is performed as in Figure 6. For ai = e-bi (b is an arbitrary positive real number), and n = 2, exponential smoothing is applicable.

Since it is difficult to handle infinity in actual computation, an asymptotic value for X(t) is obtained by sequential calculation, such as X(t) = (1 - e-b)Y(t) e-bX(t - T). Selecting n and ai appropriately also allows for averaging processing that achieves faster convergence, such as a FIR (Finite Impulse Response) filter.

Equation (1) looks apparently like an equation to simply obtain the time average of Y(t) by numerical calculation. In normal time averaging, a value sufficiently smaller than the variation period of Y(t) is selected as T, whereas in this embodiment, it is important to select, as T, a value based on the adsorption and desorption period of a PSA process. This means, instead of simply temporally averaging the measured Y(t) values, using the waveform of Y(t) during the period T as a unit to calculate the average of the waveforms obtained during a duration of an integer multiple of T preceding a certain reference time. Calculation of the average of the waveforms allows to accurately estimate a PSA pressure having characteristics that appear periodically in an interval equal to T.

In this example, a method for measuring pressure as the unit of measurement of respiratory rate has been exemplified, but since the flow rate through the tube also changes in response to pressure variations in practice, a method for measuring flow rate instead of pressure may also be used. Even a method for measuring a pressure value and a flow rate value in combination or selectively may be used, according to the operating conditions of the device or the environmental conditions.

Respiratory information obtained by arithmetic processing is so-called raw data, which represents real-time information of respiration such as pressure or flow rate in the form of a waveform. These data can be recorded or transmitted directly, but in this case, the amount of data is huge, and it takes a long time for analysis. Therefore, it is desired to automatically calculate the respiratory rate data on the respiratory rate measuring device 4, and then record and/or transmit it.

As exemplified above, a detectable respiratory waveform contains a large number of noise components generated, for example, by airflow-related pressure variations or cannula oscillation. Respiratory period and respiratory rate may not be appropriately calculated either by peak detection to estimate the times at which the pressure changes from a decrease to an increase, or by a method to detect the times at which a value equal to or greater than a threshold has been reached.

Referring to, for example, Fig. 6, the respiratory waveforms to be counted appear as peaks indicated by a1, a2 and a3 in Fig. 6 and the valleys appear immediately after these peaks, with the remaining portions corresponding to the noise components. Even when attempting to detect these peaks as the times when the pressure changes from an increase to a decrease, a portion indicated by b in Fig. 6 is very likely to be detected.

<Respiratory Rate Estimation Processing>

Even with a method for determining a portion having a value equal to or greater than a certain threshold as a peak, when the threshold is set to X, a peak as low as that indicated by a2 may not be detected, and when the threshold is set to Y, the noise component indicated by b may be detected as a peak. In this example, a waveform corresponding to only about three breaths is illustrated, but in an actual waveform, the levels of peaks or noise components may vary more seriously than those illustrated herein, and therefore it is very difficult to reliably identify peaks and set a threshold that enables reliable noise removal.

The inventors of the present invention conducted strict examination, and found that the respiratory rate can be detected with high detection performance not by the above-mentioned method, but by calculating an autocorrelation coefficient between an original waveform and a waveform shifted from the original waveform at a time At, and calculating At, at which the autocorrelation coefficient takes a peak, while At changes. These inventors also found a respiratory rate estimation method for estimating respiratory rate per predetermined time from the patient's respiratory information.

An autocorrelation coefficient R can be calculated as the following equation:

[Ec. 2]

where At is the amount of time shift, to is the data acquisition interval, n is the number of data used for one autocorrelation coefficient calculation operation, and f(t) is the value of the patient's respiratory information data at time t, which is acquired n times in the interval to. In the calculation, p and a are the average and standard deviation of f(t), but in the actual calculation, the average and standard deviation of the n obtained values of f(t) may be used. The method for calculating the autocorrelation coefficient in this embodiment is merely an example, and other methods may be used to calculate the autocorrelation coefficient.

The autocorrelation coefficient R calculated in this embodiment is multiplied by a normalization coefficient expressed as:

[Ec. 3]

1

(n -A t)a ¿

set its value in the range -1.0 to 1.0.

Figure 8 is a graph generated by plotting, with respect to At, the calculation result of the autocorrelation coefficient between the calculated waveform illustrated in Figure 6 and the waveform obtained by shifting the original waveform along the time axis by the time At, using the above-mentioned equation multiplied by the normalization coefficient. The autocorrelation coefficient is calculated every time Lt increases in t from At = 0. When At = 0, the autocorrelation coefficient is 1 because it represents the correlation between the original waveforms. After that, points P1 and P2 are obtained as points having a high correlation.

Fig. 8 reveals that even for a waveform that can hardly be used to determine the respiratory period based on a threshold or by peak detection in raw patient respiratory information data, the respiratory period can be easily estimated by calculating the autocorrelation coefficient.

The data interval of f(t) used to calculate the autocorrelation coefficient is determined from the interval of the respiratory period to be calculated. The inventors of the present invention conducted a comprehensive examination and found that the data interval of f(t) may preferably be set to two or more times of a minimum respiratory period to be calculated. The respiratory rate of an average adult is about 12 to 20 breaths per minute, which corresponds to a respiratory period of about 3 to 5 seconds. These inventors thus concluded that the data interval of f(t) may preferably be set to 5 seconds x 2 = 10 seconds.

It was found that the autocorrelation coefficient threshold for determining a peak of the autocorrelation coefficient may preferably be set to at least 0.3, and more preferably set in the range of 0.3 to 0.7. It was also found that setting the threshold below 0.3 increases the likelihood that an accidental increase in the autocorrelation value due to, for example, noise or baseline instability will be mistakenly determined as a peak, and setting the threshold above 0.7 increases the likelihood that an increase in the autocorrelation value corresponding to the respiratory period will be missed.

Respiratory rate is calculated as the reciprocal of the respiratory interval which is the value of At when the autocorrelation coefficient takes a certain value or more and a peak value. Lt at point P1 which is the first peak in Fig. 8 is determined as the respiratory interval, and dividing one minute by the respiratory interval yields a respiratory rate per minute. Respiratory rate or the number of breaths per minute (bpm) is given by the following equation:

Respiratory rate per minute (bpm) = 60 seconds / (number of steps at peak point x sampling time) Depending on the respiratory waveform, a plurality of points may be present as peaks having values equal to or greater than a certain value, as illustrated in Figure 8, but the right peak P2 is a multiple period peak appearing when At is shifted by several breaths. In estimating the respiratory rate, the leftmost peak P1, i.e. having the smallest At value, which is estimated to correspond to the basic breathing period, can be preferably used.

The calculation described in this embodiment makes it possible to automatically and accurately calculate respiratory rate information from the respiratory waveform obtained by the respiratory rate measuring unit connected to the concentrator.

When the patient respiratory information data f(t) representing a change in information related to, for example, pressure, based on exhalation and inhalation, contains only a small noise other than breathing, the autocorrelation coefficient may accidentally become high with a certain period. In view of this, by additionally performing the determination based on the variance of the patient respiratory information data f(t), the estimation unit 723 can achieve higher accuracy by calculating the absence of respiratory rate when no waveform having a magnitude high enough to determine that breathing is expected to be included is available. The determination of the presence or absence of breathing is not limited to the determination based on the variance of the patient respiratory information data f(t), and may be performed using other determination methods.

A variance a2 of the patient respiratory information data f(t) is calculated as the following equation:

[Ec. 4]

where n is the number of data used for one autocorrelation coefficient calculation operation, f(t) is the value of the patient's respiratory information data at time t, and p and a2 are the mean and variance of f(t).

As an example, the threshold for assessing variance in the use of the PSA oxygen concentration device is set as follows:

When the respiratory rate is less than eight breaths: a2 > 18 (Pa2)

When the respiratory rate is eight to 10 breaths: a2 > 18 (Pa2)

When the respiratory rate is 11 breaths or more: a2 > 7 (Pa2)

The value of this threshold, however, may vary depending on the oxygen delivery device used. Because of its variations depending on parameters such as respiratory rate, the threshold for variance assessment may be set appropriately according to the oxygen delivery device used.

As an example, a first threshold TH<d1> for the variance a2 - of the patient's respiratory information data f(t) when the respiratory rate is less than eight breaths is set to THd<1>= 18. A second threshold THd<2> for the variance a2 - of the patient's respiratory information data f(t) when the respiratory rate is eight to 10 breaths is set to TH<d2>= 18. A third threshold TH<d>3 for the variance a2 - of the patient's respiratory information data f(t) when the respiratory rate is 11 breaths or more is set to THd3 = 7. When the patient's respiratory information data f(t) is equal to or greater than the variance a2 threshold, the presence or absence of respiration is determined.

Fig. 9 is a flowchart illustrating an exemplary processing of patient respiratory information data extraction.

The patient's respiratory information data extraction processing illustrated in Fig. 9 is performed by the microcomputer unit 7 according to the computer program stored in the storage unit 71 in advance. The detected data acquisition unit 721 acquires pressure data Y(t) from the pressure sensor 6 serving as the detection unit (ST101). The arithmetic operation unit 722 calculates an average X(t) of Y(t) corresponding to n periods (ST102). The arithmetic operation unit 722 extracts patient's respiratory information data f(t) by calculating the difference between Y(t) and X(t) (ST103). When using PSA pressure variations measured and stored in advance, the arithmetic operation unit 722 does not perform the ST102 process, and extracts data f(t) of patient respiratory information by calculating the difference between Y(t) and the PSA pressure variations measured and stored in advance, such as ST 103, after the ST101 process.

During the operation of the respiratory rate measuring device 4, the respiratory rate estimation processing is performed repeatedly, so that the patient's respiratory information data f(t) is extracted and updated each time, for example, n pieces of measurement data of the pressures Y(t) are acquired at the interval to. The patient's respiratory information data f(t) may even be extracted at a predetermined interval and updated.

Fig. 10 is a flowchart illustrating an exemplary processing of respiratory rate estimation, based on patient respiratory information data.

The respiratory rate estimation processing illustrated in Fig. 10 is performed by the microcomputer unit 7 according to the computer program stored in the storage unit 71 in advance. The estimation unit 723 calculates an average p and a standard deviation a of the patient's respiratory information data f(t) (ST201). First, to determine the presence or absence of a respiratory rate, the estimation unit 723 determines whether the variance a2 i.e. the square of the standard deviation a is equal to or greater than a predetermined threshold TH<d>(ST202). When the variance a2 is less than the threshold TH<d>(NOT in ST202), the respiratory rate output unit 724 outputs information indicating that the calculation has been impossible (ST213), and the process ends. When the variance a2 is equal to or greater than the threshold TH<d> (YES in ST202), the estimation unit 723 sets the time At to zero (0) as the amount of shift in time (ST203).

The estimation unit 723 increments the time At by the data acquisition interval to (ST204). The estimation unit 723 calculates an autocorrelation coefficient R(At) of the patient respiratory information data f(t) during the time At (ST205). The estimation unit 723 writes the calculated autocorrelation coefficient R(At) to the storage unit 71 (ST206). The estimation unit 723 determines whether the time At has reached nto (ST207). When the time At has not reached nto, the process returns to ST203, where the estimation unit 723 repeats a series of processes (NOT in ST207).

When time At has reached nto (YES in ST207), the estimation unit 723 reads the autocorrelation coefficients R(At) stored in the storage unit 71 (ST208). The estimation unit 723 compares the read autocorrelation coefficients R(At) with each other to determine whether a maximum autocorrelation coefficient Max(R(At)), i.e., a peak autocorrelation coefficient R(At) (ST209), is present. More specifically, the estimation unit 723 determines whether Max(R(At)) is equal to or greater than a predetermined threshold THc.

When the maximum autocorrelation coefficient Max(R(At)) is equal to or greater than the predetermined threshold TH<c>(YES in ST209), the estimation unit 723 sets the time At for the maximum autocorrelation coefficient Max(R(At)) as a respiratory interval (ST210). The estimation unit 723 estimates the respiratory rate from the respiratory interval At (ST211). The respiratory rate output unit 724 outputs a respiratory rate signal (ST212) and ends the process. For example, the display unit 8, after receiving the respiratory rate signal, displays the respiratory rate.

When the maximum autocorrelation coefficient Max(R(At)) is less than the predetermined threshold TH<c>(NO in ST209), the respiratory rate output unit 724 outputs information indicating that the calculation has been impossible (ST213) and ends the process. During the operation of the respiratory rate measuring device 4, the respiratory rate estimation processing is repeatedly performed, so that the respiratory rate measuring device 4 can update the respiratory rate and display it on the display unit 8. The display on the display unit 8 may be performed at a predetermined interval.

LIST OF REFERENCE SIGNS

1 PSA oxygen concentration device

2 Tube

3 Nasal cannula

4 Respiratory rate measuring device

5 Hole

6 Pressure sensor

7 Microcomputer unit

71 Storage unit

72 Processing unit

721 Data acquisition unit detected

2 Arithmetic operation unit

3 Estimation unit

4 Respiratory rate output unit

Display unit

Claims (6)

1. Respiratory rate measuring device (4) comprising: a detection unit (6) adapted to detect a tube pressure and/or a tube gas flow rate in a tube (2) that supplies concentrated oxygen gas to a patient from a pressure swing adsorption oxygen concentration device (1) connected to the patient and that concentrates oxygen in the air by periodically repeating pressurization and depressurization, and output pressure data and/or gas flow rate data; an arithmetic operation unit (722) adapted to extract respiratory information data of the patient based on the pressure data and/or the gas flow rate data; and an estimation unit (723) adapted to estimate a respiratory rate per predetermined time, based on the patient's respiratory information data, characterized by The arithmetic operation unit (722) is adapted to estimate variation value data representing a periodic pressure change and/or a flow rate change of the concentrated oxygen gas occurring after an operation of an oxygen concentration device, based on the pressure data and/or the gas flow rate data, and extract the respiratory information data of the patient by removing the variation value data from the pressure data and/or the gas flow rate data, and because The estimation unit (723) is adapted to calculate an autocorrelation coefficient between the patient's respiratory information data and the shifted data of the patient's respiratory information data at a time At, while changing the time At, and estimate the respiratory rate using, as a respiratory interval, a time At at which the autocorrelation coefficient takes a peak. 2. The respiratory rate measuring device (4) according to claim 1, wherein the estimation unit (723) is adapted to estimate the respiratory rate, using a waveform indicating respiratory information data of the patient of at least 10 seconds. 3. The respiratory rate measuring device according to claim 1 or 2, wherein the estimation unit (723) is adapted to use a peak of the autocorrelation coefficient exceeding a predetermined threshold to estimate the respiratory rate. 4. The respiratory rate measuring device according to claim 3, wherein the threshold takes a value of 0.3 to 0.7. 5. The respiratory rate measuring device (4) according to claim 3 or 4, wherein the information indicating that the calculation has been impossible is output when a peak of the autocorrelation coefficient exceeding the predetermined threshold is not obtained. 6. The respiratory rate measuring device (4) according to any one of claims 1 to 5, further comprising a storage unit (71) adapted to store the variation value data measured in advance, wherein the arithmetic operation unit (722) is adapted to extract the respiratory information data of the patient by removing the variation value data from the pressure data and/or the gas flow rate data.